Tuesday, July 28, 2009

Is current widespread use of Tamiflu damaging our ability to fight a serious Influenza virus in the future?

With Tamiflu in danger of causing more deaths than Swine Flu, and the European Medicines Agency extending its shelf life to prevent embarrasing disposals by European governments, the current outbreak of H1N1 is providential to say the least. Is the use of Tamiflu to control an essentially harmless illness likely to cause future problems in treating more serious illnesses? The investigation continues.

The widespread use of antivirals increases the risk of the virus mutating and developing a resistance to the treatment. With Tamiflu, during clinical trials it was noted that the resistance rate was 0.33% in adults, 4.0% in children and 1.26% overall. It should be noted, however, that clinical trials are limited in number and, just as Adverse Drug Reactions (ADRs) are not necessarily revealed across a limited sample, the same applies to viral resistance. At the clinical trial stage, it was noted that because Tamiflu is a neuraminidase inhibitor, it would block viral replication and therefore it was unlikely that strains of Influenza would develop immunity to a significant degree. It was confidently asserted that Tamiflu would prove more effective, and that virii would be less resistant to it than to the older anti-viral treatments, amantadine and rimantadine. Where resistance developed during the clinical trials, it involved a mutation of the virus which also weakened the ability of the virus to infect humans.

That was the theory. In practise, it was rather different. As early as 2005, during the H5N1 Avian flu outbreak it was discovered that Tamiflu resistant strains were becoming increasingly prevalent. Investigations in Vietnam discovered 2 cases of tamiflu resistance out of 8 cases studied, including at least one case where H5N1 was detected at an early stage, and Tamiflu treatment was completed. The problems appeared to be caused by dosage levels: while the recommended dosage was 2 x 75mg/day, at this level viral replication was not halted completely: indeed, it required 2 x 150mg/day to completely halt replication. This continued replication during treatment meant that in some instances resistance could develop during treatment. Both of the two resistant cases studied proved fatal to the patients. While during clinical trials it was found that a 300mg dosage level per day caused no increase in ADRs, it should be borne in mind that as discussed in the previous article, neuropsychiatric damage was not recognised as an ADR until the drug was widely used.

The 2005 study looked at resistance developed by the H5N1 strain of Influenza A. By 2007, resistance was developing in seasonal flu strains of the H1N1 strain to which Swine Flu belongs. The Centre for Disease Control in Atlanta warned in 2008 that Tamiflu may prove ineffective against Influenza A flu types, and studies into seasonal strains over the 2007-2008 flu season showed that resistance of H1N1 was running at 12.3%. this was consistent with WHO studies in Canada which showed 8 out of 81 samples were resistant to Tamiflu in the same year. By the following year, the situation had worsened considerably, and preliminary CDC data showed that H1N1 resistance had reached a staggering 98.5%. Interestingly, the report (referenced below) found that contrary to the findings of the clinical trials, Tamiflu resistant strains were no less virulent than non-resistant strains. Following this, in June of this year, Nature Biotechnology magazine published an article recommending the stockpiling of alternative treatments in the event that the Swine Flu strain developed resistance. Such warnings were met with varying degrees of action by governments who had invested heavily in Tamiflu stockpiles: studies showed that, due to lower usage, resistance rates to Relenza were markedly lower. Similar results were demonstrated in studies in Europe, and there as in Canada, Relenza remained effective.

Notwithstanding this, on the 22nd July this year, CBC News in Canada reported one case of Tamiflu resistant Swine Flu in a 60 year old man from Quebec, and noted 4 other cases had so far been reported in Denmark, Japan and Hong Kong. While there was no evidence to suggest that the Canadian patient had passed the resistant virus to anyone else, it was an entirely predictable turn of events.

So where does this leave us? The current strain of H1N1 causing Swine Flu is responding to Tamiflu treatment at the current time, although the emergence of a resistant strain is a matter of concern, or at least it would be had Swine Flu proved as serious as governments would have us believe. More to the point, the widespread use of self-prescribed Tamiflu in self-diagnosed cases of Swine Flu in the UK does increase seriously the chances of a more widespread propagation of Tamiflu resistant Swine Flu, particularly as the dosage levels are below those required to completely halt viral replication.

More importantly, there is a longer term trend here. Since the introduction of Tamiflu, and more rapidly since its widescale deployment to fight H5N1 in 2005, there has been increased resistance across all virus stains of Influenza A, while Japanese researchers also found resistance in Influenza B strains amongst those who had not previously been treated with Tamiflu at a prevalence of 1.7%.

The acid test will be the rate at which the mutated H1N1 Swine Flu tamiflu resistant strain spreads. As of 2 days ago, there were 5 cases: what rate will that reach over time, and what wider impact will it have on the effectiveness of Tamiflu as an anti-influenza medicine? There is an alternative in the form of the much less widely used Relenza, but that alternative is both expensive and difficult to administer: it can not be taken in either capsule or suspension form, only via an inhaler. The cost issue is, of course, one of the reasons why Tamiflu has to date been the preferred treatment.

In terms of the widespread (a) stockpiling and (b) current use of Tamiflu, the problems of increased resistance caused by essentially unregulated usage raises more questions. Are the 5 cases of resistance a true reflection of the level of resistance acquired by the H1N1 Swine Flu strain? It is almost impossible to tell, because the symptoms of Swine Flu in many of those infected are so mild as to be almost non-existent, but if the 'pandemic' continues at its current rate, and if viral resistance increases as it did among H1N1 seasonal flu virii over a 12 month period in United States, it is possible that the widespread early use of Tamiflu as a prophylaxis will, rather than combatting the illness, contribute to an equally virulent strain which can be treated effectively only with Relenza. The British governments last purchase of anti-viral medicines included only 20% Relenza to 80% Tamiflu, and yet the British government is creating the ideal conditions for the mutation of the strain into a resistant one.

References:

National Centre for Biotechnology Information: Oseltamivir (Tamiflu) and its potential for use in a pandemic:
http://www.ncbi.nlm.nih.gov/pubmed/15709056

New England Journal of Medicine: Oseltamivir resistance during treatment of Influenze A (H5N1), Dec22 2005, De Jong, Thanh and others:
http://content.nejm.org/cgi/content/full/353/25/2667

Got the flu? Tamiflu may not be much help, says the CDC:
http://www.physorg.com/news148968276.html

Journal of the American Medical Association: Infections with Oseltamivir resistant Influenza A (H1N1) Vrus in the United States, Dharan, Gubareva, Meyer et al
http://jama.ama-assn.org/cgi/reprint/301/10/1034

CBC News - Tamiflu resistant strain rare:
http://www.cbc.ca/health/story/2009/07/22/swine-flu-tamiflu-resistance.html

Sunday, July 26, 2009

Are you more likely to die of Tamiflu than of Swine Flu?

With the expiry dates on Tamiflu extended by the European Medicines Agency, the UK holds sufficient stocks for over 28 million UK residents to receive a course of treatment.

The government would argue that it has through the NHS a duty to prevent illness in the general population, and that dosing almost half the population with self-prescribed Tamiflu is a part of fulfilling this duty. Is it? As already discussed, the apparent mortality rate for swine flu is between 1 in 4,000 and 1 in 12,000. Using the governments own suggestion that up to 30% of the UK population might eventually be infected, that means that the odds of dying from swine flu are between 1 in 15,000 and 1 in 40,000 for each member of the UK population. To put this into perspective, you stand more chance of dying of electrocution (1 in 5,000), drowning (1 in 8,943) or even dying in a plane crash (1 in 20,000) over your lifetime than you do of dying of swine flu. It is only slightly more risky than your chances of dying of capital punishment in the US (1 in 58,618) or dying in a tornado (1 in 60,000).

Be that as it may, it still leaves us with potentially over 28 million people taking Tamiflu. The question must then be whether or not Tamiflu is worse than swine flu: are you more likely to die of the cure than you are to die of the disease?

In Japan, Tamiflu is used extensively, and estimates indicate that over 30 million Japanese citizens have received doses. It is recognised that as with any drug, there are Adverse Drug Reactions (ADRs) inherent in taking Tamiflu, the most common (occuring in over 1% of patients in clinical trials) being nausea, vomiting, abdominal pain, diarrheoa and headaches. Uncommon ADRs included cardiac arrythmia, hepatitis, skin rashes including Stevens-Johnson syndrome and allergic reactions. All of these, or at least most of them, are listed on the labelling indications accompanying doses of Tamiflu sold in the European Union, as directed by the European Medicines Agency (EMEA).

However, an increasing body of medical evidence has raised concerns about other ADRs, including severe neuropsychiatric reactions, including self-harm, hallucinations, abnormal behaviour, impaired consciousness and, in infants and babies, Sudden Infant Death Syndrome, or Cot Death as it is more commonly known.

These reactions were most prominent in young adults, adolescents and children, and in fact so seriously was it taken by the Japanese Ministry of Health, Labour and Welfare (MHLW) that in 2007 manufacturers were required to alter the packaging to reflect this increased risk and to list the potential neuropsychiatric problems. The Japanese government also issued a warning that recommended Tamiflu should not be given to those aged 10 to 19 after studies showed that between 2004 and March 2007, fifteen people aged 10 to 19 have been injured or killed by jumps or fallen from buildings after taking Tamiflu, and one 17-year-old died after he jumped in front of a truck. A renewed investigation of the Japanese data was completed in April 2007. It found that 128 patients had been reported to behave abnormally after taking Tamiflu since 2001. Forty-three of them were under 10 years old, 57 patients were aged 10 to 19, and 28 patients were aged 20 or over. Eight people, including five teens and three adults, had died from these actions.

To determine whether to lift the 2007 ban, a research team from the Japanese MHLW studied 10,000 children under the age of 18 who had been diagnosed with influenza since 2006. The study was finalised in April 2009. Taking into account all degrees of abnormal behaviour, including minor behavioural problems such as incoherent speech, the study found that children who took Tamiflu were 54 per cent more likely to exhibit abnormal behaviour than those who did not take the drug. When the team limited its analysis to children who had displayed serious abnormal behaviour that led to injury or death, it found those who had taken Tamiflu were 25 per cent more likely to behave unusually.

So seriously were these findings taken that in November 2006, the US Food & Drug Administration (FDA) amended the warning label to include the possible side effects of delirium, hallucinations, or other related behavior. This went further than the FDA's previous pronouncement, from a year before, that there was insufficient evidence to claim a causal link between Tamiflu use and the deaths of 12 Japanese children (only two were from neurological problems, although more have died since then). The change to a more cautionary stance was attributed to 103 new reports that the FDA received of delirium, hallucinations and other unusual psychiatric behavior, mostly involving Japanese patients, received between August 29, 2005 and July 6, 2006. This was an increase from the 126 similar cases logged between the drug's approval in 1999 and August 2005. It should be noted that the increase of 103 fresh reports was likely to be because of greater understanding of potential neuropsychiatric ADRs, and could raise considerably the incidence of reported ADRs of this type, particularly since the 2007 studies conducted by the Japanese MHLW.

The US FDA report largely concurred with the findings of the Japanese MHLW, noting that from the date of marketing authorisation of tamiflu until 31st May 2007, it had recorded 2004 serious (ie life threatening or otherwise medically significant) ADRs out of a total of 2064 ADR reports, including 157 deaths, 43 in the US alone. More significantly, in the period 23rd April 2005 to 31st May 2007, it had recorded 939 serious ADRs worldwide, of which over 43% were in paediatric cases involving patients under 16, leading to 15 deaths out of a total of 82 recorded for all ages. Statistical analysis showed the average age of those paediatric cases as a whole was just under 9 years, while if neuropsychiatric ADRs were excluded, the average age fell to 5 years although the death rate was much higher: out of 112 cases, there were 12 deaths reported - all described as sudden death - an almost 11% mortality rate for those infants who suffered severe non-neuropsychiatric ADRs.

So what was the EU's EMEA doing during all this? In February 2007, it warned that patients receiving Tamiflu should be 'closely observed' and did move to include warnings on documentation accompanying the drug that neuropsychiatric problems could occur. However, it stopped short of restricting the prescribing of Tamiflu to young people, deciding instead that 'the benefits outweigh the risks' and that it would 'monitor closely' continued research into reports of abnormal behaviour.

With the inception of the governments self-prescription service via a telephone hotline, fresh questions have arisen. I phoned the helpline myself and was prescribed Tamiflu for myself and my children, and despite asking specifically about any dangers was re-assured that Tamiflu was 'perfectly safe', an assertion which flies in the face of studies and medical evidence. Dr Stephen Maxwell, a senior lecturer at the Clinical Pharmacology Unit of the University of Edinburgh, in an editorial in the British Medical Journal in 2007, noted that since the impact of the common flu is usually so modest in otherwise healthy people, patients should be encouraged to use instead "conservative strategies such as resting, increasing fluid intake, and taking simple analgesics and over the counter symptomatic remedies", and warned that serious side effects of a drug sometimes only come to light after the drug is launched.

"Although common adverse effects of a drug may emerge in prelicensing studies, the detection of rarer and potentially more serious events has to await exposure of large numbers of patients."

In response to Dr Maxwells editorial, Rokuro Hama, the Chairman of the Japan Insitute of Pharmacovigilance, wrote to the BMJ citing in some detail the Japanese findings. Researchers there had discovered 1,377 cases of ADRs by May 31st 2007, of which 567 were serious neuropsychiatric cases, 211 of which involved abnormal behaviour. There were also 80 reported deaths, of which 18 were sudden deaths in those aged under 10: some of these had apparently yet to reach the FDA when it conducted its paediatric review into Tamiflu ADRs, hence the slight discrepency in numbers. Most of these deaths were in infants and resembled Sudden Infant Death Syndrome, and research on young rats suggested that it was caused by a build up of Tamiflu in the brain because the Brain Blood Barrier (BBB) is not sufficiently developed in babies and young infants. From research findings already published at the time of writing, he suggested that the spectrum of action and toxicity of oseltamivir is very similar to that of central nervous system suppressants such as benzodiazepines, barbiturates and general anesthetics.

The government is faced with a considerable problem politically - it is stuck with millions of doses of date expired Tamiflu. The current hysteria surrounding swine flu provides an excellent opportunity to rid itself of these stocks, worth almost a quarter of a billion pounds, without having to admit it is throwing them away. However, by using Tamiflu against a relatively innocuous virus in the form of swine flu without fully considering the potential outcomes in terms of deaths caused by ADRs, is it putting the population at a greater risk than the miniscule risk of dying of swine flu? It is not possible to quote the odds of dying of Tamiflu, because (a) no-one knows how many people have taken it in total and (b) as knowledge of the potential ADRs becomes more widely known, more cases of ADRs are reported: the US FDA reports show graphically the exponential increase in reported ADRs over time.

It is possible to draw several conclusions. The first relates to the monitoring of patients, particulary adolescents and young adults, for neuropsychiatric effects. If my personal experience of the Swine Flu hotline is replicated elsewhere, then the government is allowing the prescribing of Tamiflu without making patients aware of the possible side effects or the need to monitor behavioural patterns for abnormalities. Without adequate medical supervision, this is clearly absurd, and could statistically place the most at risk groups at a higher risk of death or serious illness from Tamiflu than they face, statistically, from swine flu.

Secondly, the authorisation for Tamiflu to be used in 'pandemic' circumstances, issued by the World Health Organisation in June, allowed the EMEA to remove restrictions on the prescribing of Tamiflu to babies under 1 year of age. There exists a body of evidence to suggest that because of insufficient maturity of the BBB in the very young, those babies dosed with Tamiflu run a significantly higher risk of dying of Sudden Infant Death Syndrome or suffering respiratory distress than they do of dying of swine flu.

Thirdly, the suggestion that Tamiflu acts upon the brain in the same fashion as benzodiazepans, barbituates and general anaesthetics must surely give cause for concern to parents: would you really countenance giving your baby Diazepan because he or she had flu?

My own opinion, for what it's worth, is this. The government is saving itself a political problem regarding the wasted £250m it has so far spent on Tamiflu, but is probably risking the lives of more UK residents through ADRs than are at risk of death from Swine Flu. We must all make such decisions for ourselves, but neither I nor my children will be taking Tamiflu any time soon. Of course, to a certain extent, what the British government thinks is immaterial: even if it was minded to, it could not suspend or withdraw Tamiflu, nor withdraw its market authorisation in the UK, as such power resides at a European level, and only the European Commission, on the recommendation of the European Medicines Agency, could authorise such an action. Indeed, were the British government to act to protect its citizens, it could face action for disrupting the 'free market' in pharmaceutical products. Can we expect to see the Commission take action? Hardly. Governments across the EU are in the same position as the UK, sitting on huge stockpiles of date expiring antiviral drugs purchased at the behest of the European Commission in the first place. Which will volunteer to be the first to tell its citizens that it is throwing away £200m + because they are useless and potentially dangerous? Exactly.


Reference Information

US FDA Briefing Document on Paediatric ADRs to Tamiflu:
http://www.fda.gov/ohrms/dockets/ac/07/briefing/2007-4325b2_17_Tamiflu%20Pediatric%20Review%202007.pdf

EMEA Tamiflu Warning:
http://www.fdanews.com/newsletter/article?issueId=10018&articleId=91564

Dr Simon Maxwell, Tamiflu and neuropsychiatric problems in adolescents, British Medical Journal:
http://www.bmj.com/cgi/content/short/334/7606/1232

Rokuro Hama, response to editorial above:
http://www.bmj.com/cgi/eletters/334/7606/1232#169321

CTV Canada, repeating Dr Maxwells' comments for those without BMJ subscription:
http://www.ctv.ca/servlet/ArticleNews/story/CTVNews/20070615/tamiflu_070615/20070615?hub=Health

International Journal of Risk & Safety in Medicine - Fatal neuropsychiatric reactions to oseltamivir: Case series and overview of causal relationships, Rokuro Hama:
http://npojip.org/sokuho/published-paperJRS431.pdf

The Odds of Dying:
http://www.livescience.com/environment/050106_odds_of_dying.html

Does the use of Tamiflu decrease the chance of sucessfully using it during a future epidemic of something that is actually dangerous? This and other questions answered later today or tomorrow, depending on time! Please subscribe to this blog for automatic updates!

Saturday, July 25, 2009

Swine Flu - A beneficial crisis

Swine Flu - the beneficial crisis
With swine flu hysteria in full flow, we've seen a strange consensus amongst western governments which would seem to fly in the face of the facts. Normal common-or-garden seasonal flu has a mortality rate of around 1 in 1,000, while swine flu at the current time has an apparent mortality rate of between 1 in 4,000 and 1 in 12,000. OK, so it's silly season for the media. Parliament is in recess, government ministers are relaxing in Tuscany, and there's very little for the phalanx of journalists remaining in Westminster to report. But is this sufficient to explain the hysteria?Not really.

There is a clue as to the reason why contained within a little publicised decision by the European Medicines Agency (EMEA), the London based EU organisation which dictates what can and cannot be used in the EUs national health services. On the 8th May, it authorised an extension to the shelf-life of Tamiflu from its current 5 years to 7 years. Although it was reported in specialist trade publications directed at the pharmaceutical industry, there was no wider announcement, and in fact this alteration, which required the approval of the European Commission, did not appear directly in the Official Journal of the European Union in its own right. Instead, it appeared as an amendment to the marketing authorisation in summary form only, along with amendments to 65 other approved medicines - in effect, it was hidden in the middle of a list of trade names, with no indications as to what action had in fact been approved. Tamiflu is the primary antiviral drug used to combat Influenza A virii, and is the trade name of oseltamivir or oseltamivir phosphate. I have used the trade name throughout as it is easier to type!

The tale, however, got even more interesting. The EMEA had authorised this extension for all doses of tamiflu manufactured since 2002. Since 2002? Doses manufactured in 2002 would have date expired in 2007, so why on earth was it still held in stock, given that it would, by the date of the EMEA decision, have been 2 years out of date?

The answer can be found in the governments purchasing of antiviral stockpiles following the 2005 H5N1 Avian Flu scare. By late 2005, the British government had 14.6 million treatment doses (at a cost of around £238m, according to NICE costings of £16.36 per dose). This stockpile had been purchased on the open market, and included stocks held by manufacturer Roche, some of it dating back to its first authorisation for the European market in 2002, and having a date expiry in 2007. This must have been the case, as Roche themselves admitted to a 12 month lead time on the preparation of Tamiflu caused by the source of its active ingredient which comes from Star Anise. While it is not possible to be precise as to what percentage of these stocks expired in which year, a clue could be contained within a further order, placed by the British government in 2007 for an additional 14.7m doses of antiviral drugs, predominantly Tamiflu. While the public reason was to extend the stockpile of antiviral drugs available for public use, it is inescapable that at least some of this was to replace date expired stocks from the 2005 purchase. The real question must be why those date expired stocks were not disposed of at the time of expiry unless the government already had some indication that a future extension to shelf life would be granted ahead of the 'clinical' inquiry by the EMEA two years later?

Was this a purely British problem? It would seem highly unlikely. The French government had built up a similar stockpile of Tamiflu in 2005, and while the German government only held 3m doses because of a lack of availability, it had purchased a considerable stockpile of Relenza, another antiviral drug. Does this mean the German government had escaped? No. Relenza was noted by the Canadian Broadcasting Corporation as having a shorter shelf life than Tamiflu and, indeed, the EMEA, on the 8th June 2009, agreed to allow a shelf life extension to Relenza along the lines of that previously granted to Tamiflu, meaning that the German governments considerable stockpile could also be used up rather than discarded.

To put this into perspective, one must look at the cost involved. The British governments 2005 stockpile, as mentioned above, cost UK taxpayers around £238m. The French governments stockpile is of a similar size and cost, while the German governments Relenza stockpile would have cost rather more, as Relenza is difficult to administer and has a higher base cost than Tamiflu. Just these three governments stockpiles are worth more than £750m and, as WHO guidelines recommend similar coverage levels of antiviral drug stockpiles, it is not unreasonable to assume that across the EU governments are sitting on date expired stockpiles worth in excess of £2.5 billion. These figures reflect only those drugs purchased in 2005, not the subsequent increase in stockpile sizes of 2007, which must contain drugs manufactured in 2005/6 and worth, in the UK, the same again: raising British government exposure to almost £500m!

I am not familiar with the allocation of resources in national health services outside of the UK except in the broadest sense, so I can not comment on whether they have the same problems with the 'postcode lottery' of available treatments which the British NHS suffers from. However, the political fallout in the UK from having to dispose of almost a quarter of a billion pounds worth of date expired medicines would have been considerable, especially to those denied cancer treatments on cost grounds: particularly as there would be further disposals required as the 2007 purchased stockpile approached its expiry dates from 2010 onwards. The solution? Keep it in stock, cross your fingers, and hope for the sort of beneficial crisis which the European Union excels in exploiting.

And then along came Swine Flu, an H1N1 strain of the Influenza A virus. By far the largest group of at-risk patients is the over 60's, and yet this group already has an in-built immunity to this strain, as it was prevalent in the 1950's. Governments who had invested heavily in now date-expired Tamiflu must have breathed an enormous sigh of relief, for although Swine Flu was less dangerous than seasonal flu, and although Tamiflu only shortens the symptoms of those already infected by a day, it provided an excuse to rid themselves of what had become an enormous white elephant without political repercussions. By stirring up the media, a not difficult task for newspapers who otherwise face a summer of newspapers to fill without political content because of parliamentary recess, a huge increase in public demand for Tamiflu can be created which will not only clear date expired government stockpiles, but also manufacturers stockpiles. The net effect? Tamiflu can be re-ordered to replace the soon to be depleted depleted stockpiles, the political embarassment of disposing of upwards of £250m of otherwise useless medicines is avoided, and the problem is safely shuffled off 5 years into the future.

Is this the full extent of the story? No, not exactly. Leaving aside the political ramifications of the use of Tamiflu, there are other dangers inherent in widespread public use: side effects include suicidal behaviour, cyanosis and sudden infant death syndrome, amongst others. For more details on this and other associated issues, please check back here tomorrow, or subscribe to this blog for automated updates.

Reference Material:
EMEA extension of shelf life: http://www.emea.europa.eu/humandocs/PDFs/EPAR/tamiflu/28766209en.pdf

Official Journal of the European Union announcement:
http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:C:2009:146:0006:01:EN:HTML

CBC article discussing shorter shelf-life of Relenza: http://www.cbc.ca/news/background/avianflu/protection.html

NICE treatment costings for Tamiflu: http://www.nice.org.uk/nicemedia/pdf/TA158CostTemplate.xls

Details of UK stockpile: http://www.dh.gov.uk/en/AboutUs/MinistersAndDepartmentLeaders/ChiefMedicalOfficer/Features/DH_4102997

Details of stockpile size in 2005 and its manufacturing dates: http://www.apmhealtheurope.com/story.php?depsPage=1036&numero=21&ctx=4ac2fe87b3e59aed9b2a89d0543fde49 - the final paragraph quotes Roche as stating that because of manufacturing lead times, stocks must be anticipated at least 1 year in advance

Government announcement of further purchase in 2007: http://www.channel4.com/news/articles/society/health/minister+unveils+flu+pandemic+plans/1081757